Life is a game, take it seriously

Archive for the ‘Paper Talk’ Category

Paper Picks: ICRA 2017

In Computer Vision, deep learning, Machine Learning, Paper Talk, Robotics on July 31, 2017 at 1:04 pm

by Li Yang Ku (Gooly)

I was at ICRA (International Conference on Robotics and Automation) in Singapore to present one of my work this June. Surprisingly, the computer vision track seems to gain a lot of interest in the robotics community. The four computer vision sessions are the most crowded ones among all the sessions that I have attended. The following are a few papers related to computer vision and deep learning that I found quite interesting.

a) Schmidt, Tanner, Richard Newcombe, and Dieter Fox. “Self-supervised visual descriptor learning for dense correspondence.”

In this work, a self-supervised learning approach is introduced for generating dense visual descriptors with convolutional neural networks. Given a set of RGB-D videos of Schmidt, the first author, wandering around, a set of training data can be automatically generated by using Kinect Fusion to track feature points between frames. A pixel-wise contrastive loss is used such that two points belong to the same model point would have similar descriptors.

Kinect Fusion cannot associate points between videos, however with just training data within the same video, the authors show that the learned descriptors of the same model point (such as the tip of the nose) are similar across videos. This can be explained by the hypothesis that with enough data, a model point trajectory will inevitably come near to the same model point trajectory in another video. By chaining these trajectories, clusters of the same model point can be separated even without labels. The figure above visualizes the learned features with colors. Note that it learns a similar mapping across videos despite with no training signal across videos.

b) Pavlakos, Georgios, Xiaowei Zhou, Aaron Chan, Konstantinos G. Derpanis, and Kostas Daniilidis. “6-dof object pose from semantic keypoints.”

In this work, semantic keypoints predicted by convolutional neural networks are combined with a deformable shape model to estimate the pose of object instances or objects of the same class. Given a single RGB image of an object, a set of class specific keypoints is first identified through a CNN that is trained on labeled feature point heat maps. A fitting problem that maps these keypoints to keypoints on the 3D model is then solved using a deformable model that captures different shape variability. The figure above shows some pretty good results on recognizing the same feature of objects of the same class.

The CNN used in this work is the stacked hourglass architecture, where two hourglass modules are stacked together. The hourglass module was introduced in the paper “Newell, Alejandro, Kaiyu Yang, and Jia Deng. Stacked hourglass networks for human pose estimation. ECCV, 2016.” An hourglass module is similar to a fully convolutional neural network but with residual modules, which the authors claim to make it more balanced between down sampling and up sampling. Stacking multiple hourglass modules allows repeated bottom up, top down inferences which improves on the state of the art performances.

c) Sung, Jaeyong, Ian Lenz, and Ashutosh Saxena. “Deep Multimodal Embedding: Manipulating Novel Objects with Point-clouds, Language and Trajectories.”

In this work, point cloud, natural language, and manipulation trajectory data are mapped to a shared embedding space using a neural network. For example, given the point cloud of an object and a set of instructions as input, the neural network should map it to a region in the embedded space that is close to the trajectory that performs such action. Instead of taking the whole point cloud as input, a segmentation process that decides which part of the object to manipulate based on the instruction is first executed. Based on this shared embedding space, the closest trajectory to where the input point cloud and language map to can be executed during test time.

In order to learn a semantically meaningful embedding space, a loss-augmented cost that considers the similarity between different types of trajectory is used. The result shows that the network put similar groups of actions such as pushing a bar and moving a cup to a nozzle close to each other in the embedding space.

d) Finn, Chelsea, and Sergey Levine. “Deep visual foresight for planning robot motion.”

In this work, a video prediction model that uses a convolutional LSTM (long short-term memory) is used to predict pixel flow transformation from the current frame to the next frame for a non-prehensile manipulation task. This model takes the input image, end-effector pose, and a future action to predict the image of the next time step. The predicted image is then fed back into the network recursively to generate the next image. This network is learned from 50000 pushing examples of hundreds of objects collected from 10 robots.

For each test, the user specifies where certain pixels on an object should move to, the robot then uses the model to determine actions that will most likely reach the target using an optimization algorithm that samples actions for several iterations. Some of the results are shown in the figure above, the first column indicates the interface where the user specifies the goal. The red markers are the starting pixel positions and the green markers of the same shape are the goal positions. Each row shows a sequence of actions taken to reach the specified target.

Looking Into Neuron Activities: Light Controlled Mice and Crystal Skulls

In brain, Neural Science, Paper Talk, Serious Stuffs on April 2, 2017 at 9:50 pm

by Li Yang Ku (Gooly)

It might feel like there aren’t that much progress in brain theories recently, we still know very little about how signals are processed in our brain. However, scientists have moved away from sticking electrical probes into cat brains and became quite creative on monitoring brain activities.

Optogenetics techniques, which was first tested in early 2000, allow researchers to activate a neuron in a live brain by light. By controlling the light that activates motor neurons in a mouse, scientists can control its movement remotely, therefore creating a “remote controlled mouse” which you might heard of in some not that popular sci-fi novels. This is achieved by taking the DNA segment of an algae that produces light sensitive proteins and insert it into a specific brain neuron of the mouse using viral vectors. When light is shed on this protein, it opens its ion channel and activates the neuron. The result is pretty cool, but not as precise as your remote control car, yet. (see video below)

Besides the Optogenetics techniques that are used to understand the function of a neuron by actively triggering it, methods for monitoring neuron activities directly have also become quite exciting, such as using genetically modified mice with brain neurons that glow when activated. These approaches that use fluorescent markers to monitor the level of calcium in the cell can be traced back to the green fluorescent proteins introduced by Chalfie etc in 1994. With fluorescent indicators that binds with calcium, researcher can actually see brain activities the first time. A lot of progress have been made on improving these markers since; in 2007 a group in Harvard introduced the “Brainbow” that can generate up to 90 different fluorescent colors. This allowed scientists to identify neuron connection a lot easier and also helped them won a few photo contests.

To better observe these fluorescent protein sensors (calcium imaging), a recent publication in 2016 further introduced the “crystal skull”, an approach that replaces the top skull of a genetically modified mouse with a curved glass. This quite fancy approach allows researchers to monitor half a million brain neuron activities of a live mouse through mounting a fluorescence macroscope on top of the crystal skull.


Chalfie, Martin. “Green fluorescent protein as a marker for gene expression.” Trends in Genetics 10.5 (1994): 151.

Madisen, Linda, et al. “Transgenic mice for intersectional targeting of neural sensors and effectors with high specificity and performance.” Neuron 85.5 (2015): 942-958.

Josh Huang, Z., and Hongkui Zeng. “Genetic approaches to neural circuits in the mouse.” Annual review of neuroscience 36 (2013): 183-215.

Kim, Tony Hyun, et al. “Long-Term Optical Access to an Estimated One Million Neurons in the Live Mouse Cortex.” Cell Reports 17.12 (2016): 3385-3394.


Generative Adversarial Nets: Your Enemy is Your Best Friend?

In Computer Vision, deep learning, Machine Learning, Paper Talk on March 20, 2017 at 7:10 pm

by Li Yang Ku (gooly)

Generating realistic images with machines was always one of the top items on my list of difficult tasks. Past attempts in the Computer Vision community were only able to get a blurry image at best. The well publicized Google Deepdream project was able to generate some interesting artsy images, however they were modified from existing images and were designed more to make you feel like on drugs then realistic. Recently (2016), a work that combines the generative adversarial network framework with convolutional neural networks (CNNs) generated some results that look surprisingly good. (A non vision person would likely not be amazed though.) This approach was quickly accepted by the community and was referenced more then 200 times in less then a year.

This work is based on an interesting concept first introduced by Goodfellow et al. in the paper “Generative Adversarial Nets” at NIPS 2014 ( The idea was to have two neural networks compete with each other. One would try to generate images as realistic as it can and the other network would try to distinguish them from real images at its best. By theory this competition will reach a global optimum where the generated image and the real image will belong to the same distribution (Could be a lot trickier in practice though). This work in 2014 got some pretty good results on digits and faces but the generated natural images are still quite blurry (see figure above).

In the more recent work “Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks” by Radford, Metz, and Chintala, convolutional neural networks and the generative adversarial net framework are successfully combined with a few techniques that help stabilize the training ( Through this approach, the generated images are sharp and surprisingly realistic at first glance. The figures above are some of the generated bedroom images. Notice that if you look closer some of them may be weird.

The authors further explored what the latent variables represents. Ideally the generator (neural network that generates image) should disentangle independent features and each latent variable should represent a meaningful concept. By modifying these variables, images that have different characteristics can be generated. Note that these latent variables are what given to the neural network that generates images and is randomly sampled from a uniform distribution in the previous examples. In the figure above is an example where the authors show that the latent variables do represent meaningful concepts through arithmetic operations. If you subtract the average latent variables of men without glasses from the average latent variables of men with glasses and add the average latent variables of women without glasses, you obtain a latent variable that result in women with glasses when passed through the generator. This process identifies the latent variables that represent glasses.




The most cited papers in computer vision and deep learning

In Computer Vision, deep learning, Paper Talk on June 19, 2016 at 1:18 pm

by Li Yang Ku (Gooly)

paper citation

In 2012 I started a list on the most cited papers in the field of computer vision. I try to keep the list focus on researches that relate to understanding this visual world and avoid image processing, survey, and pure statistic works. However, the computer vision world have changed a lot since 2012 when deep learning techniques started a trend in the field and outperformed traditional approaches on many computer vision benchmarks. No matter if this trend on deep learning lasts long or not I think these techniques deserve their own list.

As I mentioned in the previous post, it’s not always the case that a paper cited more contributes more to the field. However, a highly cited paper usually indicates that something interesting have been discovered. The following are the papers to my knowledge being cited the most in Computer Vision and Deep Learning (note that it is “and” not “or”). If you want a certain paper listed here, just comment below.

Cited by 5518

Imagenet classification with deep convolutional neural networks

A Krizhevsky, I Sutskever, GE Hinton, 2012

Cited by 1868

Caffe: Convolutional architecture for fast feature embedding

Y Jia, E Shelhamer, J Donahue, S Karayev…, 2014

Cited by 1681

Backpropagation applied to handwritten zip code recognition

Y LeCun, B Boser, JS Denker, D Henderson…, 1989

Cited by 1516

Rich feature hierarchies for accurate object detection and semantic segmentation

R Girshick, J Donahue, T Darrell…, 2014

Cited by 1405

Very deep convolutional networks for large-scale image recognition

K Simonyan, A Zisserman, 2014

Cited by 1169

Improving neural networks by preventing co-adaptation of feature detectors

GE Hinton, N Srivastava, A Krizhevsky…, 2012

Cited by 1160

Going deeper with convolutions

C Szegedy, W Liu, Y Jia, P Sermanet…, 2015

Cited by 977

Handwritten digit recognition with a back-propagation network

BB Le Cun, JS Denker, D Henderson…, 1990

Cited by 907

Visualizing and understanding convolutional networks

MD Zeiler, R Fergus, 2014

Cited by 839

Dropout: a simple way to prevent neural networks from overfitting

N Srivastava, GE Hinton, A Krizhevsky…, 2014

Cited by 839

Overfeat: Integrated recognition, localization and detection using convolutional networks

P Sermanet, D Eigen, X Zhang, M Mathieu…, 2013

Cited by 818

Learning multiple layers of features from tiny images

A Krizhevsky, G Hinton, 2009

Cited by 718

DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition

J Donahue, Y Jia, O Vinyals, J Hoffman, N Zhang…, 2014

Cited by 691

Deepface: Closing the gap to human-level performance in face verification

Y Taigman, M Yang, MA Ranzato…, 2014

Cited by 679

Deep Boltzmann Machines

R Salakhutdinov, GE Hinton, 2009

Cited by 670

Convolutional networks for images, speech, and time series

Y LeCun, Y Bengio, 1995

Cited by 570

CNN features off-the-shelf: an astounding baseline for recognition

A Sharif Razavian, H Azizpour, J Sullivan…, 2014

Cited by 549

Learning hierarchical features for scene labeling

C Farabet, C Couprie, L Najman…, 2013

Cited by 510

Fully convolutional networks for semantic segmentation

J Long, E Shelhamer, T Darrell, 2015

Cited by 469

Maxout networks

IJ Goodfellow, D Warde-Farley, M Mirza, AC Courville…, 2013

Cited by 453

Return of the devil in the details: Delving deep into convolutional nets

K Chatfield, K Simonyan, A Vedaldi…, 2014

Cited by 445

Large-scale video classification with convolutional neural networks

A Karpathy, G Toderici, S Shetty, T Leung…, 2014

Cited by 347

Deep visual-semantic alignments for generating image descriptions

A Karpathy, L Fei-Fei, 2015

Cited by 342

Delving deep into rectifiers: Surpassing human-level performance on imagenet classification

K He, X Zhang, S Ren, J Sun, 2015

Cited by 334

Learning and transferring mid-level image representations using convolutional neural networks

M Oquab, L Bottou, I Laptev, J Sivic, 2014

Cited by 333

Convolutional networks and applications in vision

Y LeCun, K Kavukcuoglu, C Farabet, 2010

Cited by 332

Learning deep features for scene recognition using places database

B Zhou, A Lapedriza, J Xiao, A Torralba…,2014

Cited by 299

Spatial pyramid pooling in deep convolutional networks for visual recognition

K He, X Zhang, S Ren, J Sun, 2014

Cited by 268

Long-term recurrent convolutional networks for visual recognition and description

J Donahue, L Anne Hendricks…, 2015

Cited by 261

Two-stream convolutional networks for action recognition in videos

K Simonyan, A Zisserman, 2014


Local Distance Learning in Object Recognition

In Computer Vision, Paper Talk on February 8, 2015 at 11:59 am

by Li Yang Ku (Gooly)

learning distance

Unsupervised clustering algorithms such as K-means are often used in computer vision as a tool for feature learning. It can be used in different stages in the visual pathway. Running K-means algorithm on a small region of pixel patches might result in finding a lot of patches with edges of different orientation while running K-means on a larger HOG feature might result in finding contours of meaningful parts of objects such as faces if your training data consists of selfies.  However, although convenient and simple as it seems, we have to keep in mind that these unsupervised clustering algorithms are all based on the assumption that a meaningful metric is provided. Without this criteria, clustering suffers from the “no right answer” problem. Whether the algorithm should group a set of images into clusters that contain objects with the same type or the same color is ambiguous and not well defined. This is especially true when your observation vectors are consists of values representing different types of properties.

distance learning

This is where Distance Learning comes into play. In the paper “Distance Metric Learning, with Application to Clustering with Side-Information” written by Eric Xing, Andrew Ng, Michael Jordan and Stuart Russell, a matrix A that represents the distance metric is learned through convex optimization using user inputs specifying grouping examples. This matrix A can either be full or diagonal. When learning a diagonal matrix, the values simply represent the weights of each feature. If the goal is to group objects with similar color, features that can represent color will have a higher weight in the matrix. This metric learning approach was shown to improve clustering on the UCI data set.

visual association

In another work “Recognition by Association via Learning Per-exemplar Distances” written by Tomasz Malisiewicz and Alexei Efros, the object recognition problem is posed as data association. A region in the image is classified by associating it with a small set of exemplars based on visual similarity. The authors suggested that the central question for recognition might not be “What is it?” but “What is it like?”. In this work, 14 different type of features under 4 categories, shape, color, texture and location are used. Unlike the single distance metric learned in the previous work, a separate distance function that specifies the weights put on these 14 different type of features is learned for each exemplar. Some exemplars like cars will not be as sensitive to color as exemplars like sky or grass, therefore having a different distance metric for each exemplar becomes advantageous in such situations. These class of work that defines separate distance metrics are called Local Distance Learning.

instance distance learning

In a more recent work “Sparse Distance Learning for Object Recognition Combining RGB and Depth Information” by Kevin Lai, Liefeng Bo, Xiaofeng Ren, and Dieter Fox, a new approach called Instance Distance Learning is introduced, which instance is referred to one single object. When classifying a view, the view to object distance is compared simultaneously to all views of an object instead of a nearest neighbor approach. Besides learning weight vectors on each feature, weights on views are also learned. In addition, a L1 regularization is used instead of a L2 regularization in the Lagrange function. This generates a sparse weight vector which has a zero term on most views. This is quite interesting in the sense that this approach finds a small subset of representative views for each instance. In fact as shown in the image below, with just 8% of the exemplar data a similar decision boundaries can be achieved. This is consistent to what I talked about in my last post; human brain doesn’t store all the possible views of an object nor does it store a 3D model of the object, instead it stores a subset of views that are representing enough to recognize the same object. This work demonstrates one possible way of finding such subset of views.

instance distance learning decision boundaries


How objects are represented in human brain? Structural description models versus Image-based models

In Computer Vision, Neural Science, Paper Talk on October 30, 2014 at 9:06 pm

by Li Yang Ku (Gooly)


A few years ago while I was still back in UCLA, Tomaso Poggio came to give a talk about the object recognition work he did with 2D templates. After the talk some student asked about whether he thought about using a 3D model to help recognizing objects from different viewpoints. “The field seems to agree that models are stored as 2D images instead of 3D models in human brain” was the short answer Tomaso replied. Since then I took it as a fact and never had a second thought of it till a few month ago when I actually need to argue against storing a 3D model to people in robotics.

70s fashion

To get the full story we have to first go back to the late 70s. The study of visual object recognition is often motivated by the problem of recognizing 3D objects while only receiving 2D patterns of light on our retina. The question was whether our object representations is more similar to abstract three-dimensional descriptions, or are they tied more closely to the two-dimensional image of an object? A commonly held solution at that time, popularized by Marr was that the goal of vision is to reconstruct 3D. In the paper “Representation and recognition of the spatial organization of three-dimensional shapes” published in 1978 Marr and Nishihara assumes that at the end of the reconstruction process, viewer centered descriptions are mapped into object centered representations. This is based on the hypothesis that object representation should be invariant over changes in the retinal image. Based on this object centered theory, Biederman introduced the recognition by component (RBC) model in 1987 which proposes that objects are represented as a collection of volumes or parts. This quite influential model explains how object recognition can be viewpoint invariant and is often referred to as a structural description model.

The structural description model or object centered theory was the dominant theory of visual object understanding around that time and it can correctly predict the view-independent recognition of familiar objects. On the other hand, the viewer centered models, which store a set of 2D images instead of one single 3D model, are usually considered implausible because of the amount of memory a system would require to store all discriminable views of many objects.


However, between late 1980’s to early 1990’s a wide variety of psychophysical and neurophysiological experiments surprisingly showed that human object recognition performance is strongly viewpoint dependent across rotation in depth. Before jumping into late 80’s I wanna first introduce some work done by Palmer, Rosch, and Chase in 1981. In their work they discovered that commonplace objects such as houses or cars can be hard or easy to recognize, depending on the attitude of the object with respect to the viewer. Subjects tended to respond quicker when the stimulus was shown from a good or canonical perspective. These observations was important in forming the viewer centered theory.

Paper clip like objects used in Bulthoff's experiments

Paper clip like objects used in Bulthoff’s experiments

In 1991 Bulthoff conducted an experiment on understanding these two theories. Subjects are shown sequences of animations where a paper clip like object is rotating. Given these sequences, the subjects have enough information to reconstruct a 3D model of the object. The subjects are then given a single image of a paper clip like object and are asked to identify whether it is the same object. Different viewing angles of the object are tested. The assumption is that if only one single complete 3D model of this object exists in our brain then recognizing it from all angles should be equally easy. However, according to Bulthoff when given every opportunity to form 3D, the subjects performed as if they have not done so.

Bulthoff 1991

In 1992 Edelman further showed that canonical perspectives arise even when all the views in question are shown equally often and the objects posses no intrinsic orientation that might lead to the advantage of some views.

Edelman 1992

Error rate from different viewpoint shown in Edelman’s experiment

In 1995 Tarr confirmed the discoveries using block like objects. Instead of showing a sequence of views of the object rotating, subjects are trained to learn how to build these block structures by manually placing them through an interface with fixed angle. The result shows that response times increased proportionally to the angular distance from the training viewpoint. With extensive practice, performance became nearly equivalent at all familiar viewpoints; however practice at familiar viewpoints did not transfer to unfamiliar viewpoints.

Tarr 1995

Based on these past observations, Logothetis, Pauls, and Poggio raised the question “If monkeys are extensively trained to identify novel 3D objects, would one find neurons in the brain that respond selectively to particular views of such object?” The results published in 1995 was clear. By conducting the same paper clip object recognition task on monkeys, they found 11.6% of the isolated neurons sampled in the IT region, which is the region that known to represent objects, responded selectively to a subset of views of one of the known target object. The response of these individual neurons decrease when the shown object rotate in all 4 axis from the canonical view which the neurons represent. The experiment also shows that these view specific neurons are scale and position invariant up to certain degree.

Logothetis 1995

Viewpoint specific neurons

These series of findings from human psychophysics and neurophysiolog research provided converging evidence for ‘image-based’ models in which objects are represented as collections of viewpoint-specific local features. A series of work in computer vision also shown that by allowing each canonical view to represent a range of images the model is no longer unfeasible. However despite a large amount of research, most of the detail mechanisms are still unknown and require further research.

Check out these papers visually in my other website

References not linked in post:

Tarr, Michael J., and Heinrich H. Bülthoff. “Image-based object recognition in man, monkey and machine.” Cognition 67.1 (1998): 1-20.

Palmeri, Thomas J., and Isabel Gauthier. “Visual object understanding.” Nature Reviews Neuroscience 5.4 (2004): 291-303.

Human vision, top down or bottom up?

In Computer Vision, Neural Science, Paper Talk on February 9, 2014 at 6:42 pm

by Li Yang Ku (Gooly)

top-down bottom-up

How our brain handles visual input is a myth. When Hubel and Wiesel discovered the Gabor filter like neuron in cat’s V1 area, several feed forward model theories appear. These models view our brain as a hierarchical classifier that extracts features layer by layer. Poggio’s papers “A feedforward architecture accounts for rapid categorization” and “Hierarchical models of object recognition in cortex” are good examples. These kind of structure are called discriminative models. Although this new type of model helped the community leap forward one step, it doesn’t solve the problem. Part of the reason is that there are ambiguities if you are only viewing part of the image locally and a feed-forward only structure can’t achieve global consistency.

Feedforward Vision

Therefore the idea that some kind of feedback model has to exist gradually emerged. Some of the early works in the computer science community had first came up with models that rely on feedback, such as Gefforey Hinton’s Boltzman Machine invented back in the 80’s which developed into the so called deep learning around late 2000. However it was only around early 2000 had David Mumford clearly addressed the importance of feedback in the paper “Hierarchical Bayesian inference in the visual cortex“.  Around the same time Wu and others had also combined feedback and feedforward models successfully on textures in the paper “Visual learning by integrating descriptive and generative methods“. Since then the computer vision community have partly embraced the idea that the brain is more like a generative model which in addition to categorizing inputs is capable of generating images. An example of human having generative skills will be drawing images out of imagination.


Slightly before David Mumford addresses the importance of the generative model. Lamme in the neuroscience community also started a series of research on the recurrent process in the vision system. His paper “The distinct modes of vision offered by feedforward and recurrent processing” published in 2000 addressed why recurrent (feedback) processing might be associated with conscious vision (recognizing object). While in the same year the paper “Competition for consciousness among visual events: the psychophysics of reentrant visual processes.” published in the field of psychology also addressed the reentrant (feedback) visual process and proposed a model where conscious vision is associated with the reentrant visual process.


While both the neuroscience and psychology field have research results that suggests a brain model that is composed of feedforward and feedback processing where the feedback mechanism is associated with conscious vision, a recent paper “Detecting meaning in RSVP at 13 ms per picture” shows that human is able to recognize high level concept of an image within 13 ms, a very short gap that won’t allow the brain to do a complete reentrant (feedback) visual process. This conflicting result could suggest that conscious vision is not the result of feedback processing or there are still missing pieces that we haven’t discover. This kind of reminds me one of Jeff Hawkins’  brain theory, which he said that solving the mystery of consciousness is like figuring out the world is round not flat, it’s easy to understand but hard to accept, and he believes that consciousness does not reside in one part of the brain but is simply the combination of all firing neuron from top to bottom.

Paper Talk: Untangling Invariant Object Recognition

In Computer Vision, Neural Science, Paper Talk on September 29, 2013 at 7:31 pm

by Gooly (Li Yang Ku)

Untangle Invariant Object Recognition

In one of my previous post I talked about the big picture of object recognition, which can be divided into two parts 1) transforming the image space 2) classifying and grouping. In this post I am gonna talk about a paper that clarifies object recognition and some of it’s pretty cool graphs explaining how our brains might transform the input image space. The paper also talked about why the idealized classification might not be what we want.

Lets start by explaining what’s a manifold.

image space manifolds

An object manifold is the set of images projected by one object in the image space. Since each image is a point in the image space and an object can project similar images with infinitely small differences, the points form a continuous surface in the image space. This continuous surface is the object’s manifold. Figure (a) above is an idealized manifold generated by a specific face. When the face is viewed from different angles the projected point move around on the continuous manifold. Although the graph is drew in 3D one should keep in mind that it is actually in a much larger dimension space. A 576000000 dimension space if consider human eyes to be 576 mega pixel. Figure (b) shows another manifold from another face, in this space the two individuals can be separated easily by a plane. Figure (c) shows another space which the two faces would be hard to separate. Note that these are ideal spaces that is possibly transformed from the original image space by our cortex. If the shapes are that simple, object recognition would be easy. However, the actual stuff we get is in Figure (d). The object manifolds from two objects are usually tangled and intersect in multiple spots. However the two image space are not the same, therefore it is possible that through some non linear operation we can transform figure (d) to something more like figure (c).

classification: manifold or point?

One interesting point this paper made is that the traditional view that there is a neuron that represents an object is probably wrong. Instead of having a grandmother cell (yes.. that’s how they called it) that represents your grandma, our brain might actually represents her with a manifold. Neurologically speaking, a manifold could be a set of neurons that have a certain firing pattern. This is related to the sparse encoding I talked about before and is consistent with Jeff Hawkins’ brain theory. (See his talk about sparse distribution representation around 17:15)

The figure (b) and (c) above are the comparison between a manifold representation and a single cell representation. What is being emphasized is that object recognition is more a task of transforming the space rather than classification.

Why Visual Illusions: Illusory Contours and Checkerboard Illusion

In Computer Vision, Paper Talk, Visual Illusion on September 16, 2013 at 6:33 pm

by Gooly (Li Yang Ku)

I talked about some visual illusions in my previous post but didn’t mention why they are important to computer vision and the pros of seeing visual illusions. In this post I am gonna talk about the advantage of having two of the most common known visual illusions, Illusory contours and checkerboard illusion.

Illusory Contours:

Kanizsa's Triangle

The Kanizsa’s triangle invented by Gaetano Kanizsa is a very good example of illusory contours. Even though the center upside down triangle doesn’t exist, you are forced to see it because of the clues given by the other parts. If you gradually cover up some of the circles and corners, at some point you would be able to see the pac man and the angle as individual objects and the illusory contours will disappear. This illusion is the side effect of how we perceive objects and shows that we see edges using various clues instead of just light differences. Because our eyes receive noisy real world inputs, illusory contour actually helps us fill in the missing contours caused by lighting, shading, or occlusion. It also explains why a bottom up vision system won’t work in many situations. In the paper “Hierarchical Bayesian inference in the visual cortex” written by Lee and Mumford, a Kanizsa’s square is used to test whether monkeys perceive illusory contours in V1. The result is positive but has a delayed response compared to V2. This suggests that information of illusory contours is possibly generated in V2 and back propagated to V1.

Checkerboard Illusion:

checker board illusion

This checkerboard illusion above is done by Edward H. Adelson. In the book “Perception as Bayesian Inference” Adelson wrote a chapter discussing how we perceive objects under different lighting conditions. In other words, how we achieve “lightness constancy”. The illusion above should be easily understandable. At first sight, In the left image square A on the checkerboard seems to be darker than square B although they actually have the same brightness. By breaking the 3D structure, the right images shows that the two squares indeed have the same brightness. We perceive A and B differently in the left image because our vision system is trying to achieve lightness invariant. In fact if the cylinder is removed square A will be darker than square B, therefore lightness constancy actually gives us the correct brightness when only constant lighting is presented. This allows us to recognize the same object even under large lighting changes, which I would argue is an important ability for survival. In the paper “Recovering reflectance and illumination in a world of painted polyhedra” by Sinha and Adelson, how we construct 3D structure from 2D drawing and shading are further discussed. Understanding object’s 3D structure is crucial in obtaining light constancy like the checkerboard illusion above. As in the image below, by removing certain types of junction clues, a 3D drawing can easily be seen flat. However, as mentioned in the paper, more complex global strategies are needed to cover all cases.

3D 2D Recovering  Reflectance  and  Illumination  in  a  World  of  Painted  Polyhedra by Pawan Sinha & Edward Adelson

I was gonna post this a few month ago but was delayed by my Los Angeles to Boston road trip (and numerous good bye parties), but I am now officially back to school in UMASS Amherst for a PhD program. Not totally settled down yet but enough to make a quick post.

Back to Basics: Sparse Coding?

In Computer Vision, Neural Science, Paper Talk on May 4, 2013 at 9:04 pm

by Gooly (Li Yang Ku)

Gabor like filters

It’s always good to go back to the reason that lured you into computer vision once in a while. Mine was to understand the brain after I astonishingly realized that computers have no intelligence while I was studying EE in undergrad. In fact if they use the translation “computer” instead of  “electrical brain” in my mother language, I would probably be better off.

Anyway, I am currently revisiting some of the first few computer vision papers I read, and to tell the truth I still learn a lot from reading stuffs I read several times before, which you can also interpret it as I never actually understood a paper.

So back to the papers,

Simoncelli, Eero P., and Bruno A. Olshausen. “Natural image statistics and neural representation.” Annual review of neuroscience 24.1 (2001): 1193-1216.

Olshausen, Bruno A., and David J. Field. “Sparse coding with an overcomplete basis set: A strategy employed by VI?.” Vision research 37.23 (1997): 3311-3326.

Olshausen, Bruno A. “Emergence of simple-cell receptive field properties by learning a sparse code for natural images.” Nature 381.6583 (1996): 607-609.

These 3 papers are essentially the same, the first two are the spin-offs of the 3rd paper published in Nature. I personally prefer the second paper for reading.

Brain Sparse Coding

In this paper, Bruno explains why overcomplete sparse coding is essential for human vision in a statistical way. The goal is to obtain a set of basis functions that can be used to regenerate an image. (basis functions are filters) This can be viewed as an image encoding problem, but instead of having an encoder that compresses the image to the minimum size, the goal is to also remain sparsity, which means only a small amount of basis are used compared to the whole basis pool. Sparsity has obvious advantage biologically, such as saving energy, but Bruno conjectured that sparsity is also essential to vision and is originated from the sparse structure in natural image.

In order to obtain this set of sparse basis, a sparsity constraint is added to the energy function for optimization. The final result is a set of basis function (image atop) that interestingly looks very similar to Gabor filters which is found in the visual cortex. This some how proves that sparseness is essential in the evolution of human vision.